Continuation in using 2018 SOC Codes
- The BLS has completed its migration of their Occupational Employment Statistics (OES) data from the 2010 to the 2018 Standard Occupation Classification (SOC) coding system, their coding system best resembles a modified version of the 2010 SOC scheme. We follow the OES occupation code scheme, with a few necessary code changes in order to maintain a complete set of detail level occupation codes.
- The O*NET-26.1 data used for the Core Competency portion of the 2020 IMPLAN Occupation Data Suite are also based on the 2018 SOC scheme.
Updated O*Net Data
The O*NET database contains a rich set of variables that describe work and worker characteristics, including education and skill requirements, for a wide array of occupations The 2020 IMPLAN Occupational Data Suite incorporates the latest version (26.1) of the O*Net data. Details about the latest O*Net data can be found on the O*Net website.
Use of ACS PUMS Experimental Data
The 2020 ACS 1-Year data normally used for IMPLAN’s synthesis of occupation data was released by the Census Bureau as an experimental set. A statistical analysis of collection methods and changes in procedure revealed that, due to the impact of COVID-19 during the collection year, the regular 1-Year estimate data was not of a sufficient quality to release as a regular set of data.
In short, data collection was disrupted due to the COVID-19 Pandemic. This resulted in a high degree of difference in social, economic, and housing characteristics between respondents and nonrespondents, which introduced nonresponse bias into the ACS data. As a result, modifications were made to the usual weighting procedure used by the ACS data. A paper describing these changes can be found here:
Though these changes are new, IMPLAN decided to utilize the 2020 ACS 1-Year data in order to produce it’s annual occupational data set for the following reasons:
- There is a lack of a suitable replacement for this data.
- Experimental weighting used in the ACS data was designed around producing estimates at the state level. The data that IMPLAN uses from this set is all at the national level, so it is presumed to be less affected.
Even though the ACS data does not form the bulk of the 2020 IMPLAN Occupational data set, users of that data should use caution when making comparisons between different occupational data set years as a result of this change to the raw data.
Written April 14, 2022